My DIWBIS final project is coming along in a rather organic way so far. I’ve continued tagging resources as they are posted in our online class environment. My intention is to review these bookmarks, refine the tagging, and add annotations before the project due date. The refining portion of this endeavor has created some uncertainty for me. When possible, I have used the poster’s own terms for tagging the bookmark. In other cases, no usable terms are associated with the link, and I am left to make a decision regarding the “aboutness” of the resource. This is often done “off the cuff.”

I am saving and organizing these resources for the purpose of discovery (and preservation but let’s save that for a later date). An assigned reading this week, Usage Patterns of Collaborative Tagging Systems, should prove useful as I begin the process of refining the tags for this purpose. The authors discuss the function that tags perform for bookmarks. If I begin to think about the way a tag will be used, it should help me tag the resource in a way that aids in its discovery. I think I’ve done this unconsciously to some extent, but the list presented in the article will allow for a more methodical approach. Of course, the list includes the resource’s “aboutness” that must be considered. The article also lists other functions that are worth consideration. For instance, “identifying what it is.” Is it an article, a blog, or video? Or, evaluating resources to determine the task involved and grouping these items accordingly. In other words, should the HTML tutorial be tagged with “web design?”

In addition to considering tags in relation to function, the authors touch upon another point that I’ve been mulling over. That is, tagging items consistently in regard to parts of speech and singular verses plural forms. Another article, from D-Lib, has some suggestions for “tidying up” tags. These include:

  • using plurals rather than singulars
  • using lower case,
  • grouping words using an underscore,
  • following tag conventions started by others and
  • adding synonyms.

Although there is some overlap, Ideant also suggests ways to ensure the value of tags in distributed classification systems.

Here’s our tag cloud now, in its pre-refinement stage. It’s a little fuzzy because I had to save the web page as an image and crop in to isolate the tag cloud…(wordpress doesn’t allow Diigo’s Javascript).